Phase One IQ3 100mp. Shipping now. Full frame CMOS. HDMI.

[Jamgolf]

Just as I can't afford a Ferrari 458 but I do admire it, I simply can't afford to buy this IQ3-100. However as a photography enthusiast I really appreciate that it has been developed and that it exists. Kudos Phase One !

I have a feeling that when a comprehensive set of tests becomes available, with various Rodenstock tech-camera lenses, the results will knock the socks off most 'actual' users. Those who can afford the back will be (and are currently) ordering IQ3-100 in significant numbers. I predict Phase One will have a killer fiscal year.

Now, 32mm is a really special lens but shifting 12mm left/right even on a 60MP back it loses most of its magic, IMHO. While those shifted corners (on 60MP back) don't have color casts, they also do not have the wow-factor. For that wow-factor, I am learning to limit the shifts to 8-9mm. I predict if IQ3-100 can yeild clean 8mm shifts with 32HR, it will still be a killer combo and those who can afford will be buyers.

I really feel the release of a full-frame CMOS digital back is a rejoicing moment for all those who share the passion of photography.

Major props to Yat Lee for posting the 32HR examples.

 

[torger]

I think it's a slippery slope. For each new back we seem to accept further reductions in movement range. When does a tech cam stop being a tech cam?

I'm not really into maximizing resolving power at all costs. I want to have compositional flexibility too. I already thought the Dalsa 80MP had a bit too much issues on the rodie wides and the 60MP is a better choice.

The 32HR is a very wide lens with the 645 though, I'd say it's primarily a speciality lens for architecture shooters. For landscape I think 40mm/645 is a more reasonable widest angle. The 40HR is less retrofocus than the 32HR though so I'd expect similar issues.

On the 40mm I'd like to have 12mm shift for 645 fullframe as minimum to get about the same flexibility I have (and use) in my current system, but it wouldn't hurt with a little bit more. The IQ3 actually might be able to do 12mm on the 40HR... we'll see.

 

[vjbelle]

There does seem to be some diminishing returns with regards to movements but I don't need anywhere near as much as some. Even 12mm shift would be overkill for me as 10mm shifting left and right results in an almost perfect 16X9 image which is my goal for horizontal shifting. For vertical rise I agree that it may be beneficial to want slightly more than 10mm but what I have found is that only with my longer lenses (120mm and beyond) do I keep the kind of image quality I want when shifting beyond 10mm. Sometimes, for me, the results are better by rotating around the nodal. On my last trip to the Dolomites I would never use rise beyond 10mm.... if I needed more height for Architectural images I would just tilt the camera a little and correct in post. The small amount of perspective correction is very easily handled and I still maintain all of the detail I expect from the given lens. I never shot anything that needed wider than my Schneider 60mm XL. So, maybe this back is a good fit for me. I'll find out soon enough as my order is in.......

Victor

 

[Ken_R]

I think it's a slippery slope. For each new back we seem to accept further reductions in movement range. When does a tech cam stop being a tech cam?

I'm not really into maximizing resolving power at all costs. I want to have compositional flexibility too.

Interesting question.

Even with limited shift capabilities on a tech cam it is more than you can have with all Medium Format SLR systems with the exception of the Hasselblad 1.5x HTS (that has a huge size/weight and $$ penalty). Also straight on the tech cam lenses are amazing and generally much better than SLR lenses. They are also smaller and more compact and easy to filter (most of them). Ok lets say some back / lens combinations do not allow almost any shift, you can always use tilt which is still quite useful in a lot of situations.

Also, a system like the ALPA FPS (can I call it a tech cam?) will allow you to get ultra wide angle coverage in medium format digital unavailable from the MF SLR systems (with the use of the Canon 17 and 24 TS-E II lenses).

So the Tech Cam has many possible advantages. Small size and weight is one that even without tilt and shift capabilities one can have if desired.

So yes, the reduced shiftability of the new sensors is a bummer and nudge people into overlooking tech cams but IMHO the tech cams still have their place.

But just in case it wouldn't hurt to stock up on some of the better tech cam lenses available now.

 

[MrSmith]

"I really feel the release of a full-frame CMOS digital back is a rejoicing moment for all those who share the passion of photography."

You drank the KoolAid :facesmack:

 

[Jamgolf]

"I really feel the release of a full-frame CMOS digital back is a rejoicing moment for all those who share the passion of photography."

You drank the KoolAid :facesmack:

Oh - you got me there :clap:

 

[vjbelle]

After seeing some of the horrible posted images I thought I would at least test my Leaf/Credo 50 to the limits. I know.... its not the 100 but its a Sony sensor and probably the basis for the 100. I used my 35mm XL and added rise of 15mm...... that's 15mm folks on the this legendary love/hate lens. Please forgive any sensor dirt..... I think that Phase will have cast issues under control. I would never, ever shift my 35mm XL 15mm but this is just an example of cast with CMOS and the Phase fix.

Victor

 

[torger]

As noted in a neighboring thread there's currently some issues with the LCC algorithm in Capture One (it's a nice tradition Phase One has ), so it's easy to get confusing results.

 

[narikin]

So the Tech Cam has many possible advantages. Small size and weight is one that even without tilt and shift capabilities one can have if desired.

and: stitching for XXL file size/resolution
and: a tech cam lens will out-resolve a MF dSLR lens

I compared the Phase One -Schneider 55mm with the Roddy 60mm, and it was very easy to see how much better the tech lens was (everywhere but the very center, where it was more evenly matched). If you want to walk around doing landscapes with some normal wide (40-60mm) you are better off with a tech lens' resolving power than one designed for your DF/XF camera.

(I don't think its that Schneider/Phase make bad lenses, but that it's a function of strong retrofocus vs mild retrofocus)

 

[ondebanks]

Having a 15-stop sensor with 14-bit ADCs will just give you really clean shadows since some of the noise will be below the limits of the ADC's ability to "see" it.

In that situation, the normal approach would be to increase the ADC gain, enough to squeeze those 15 stops into 14 bits of output - which is better than throwing away that last stop of DR. The tradeoff - just a tiny amount of quantization noise.

Ray

 

[Stefan Steib]

the problem with the Whiteshot fix is that in these areas the green channel will be about 3-5 stops underexposed compared to the rest of the image.
thus it will become incredibly noisy. CMOS has some more reserves for this than CCD, but doing any decent printfile with this, will be very difficult for post.

Oh and as resolution in these outer areas of the greenchannel will suffer as well, the 100 Mpix of resolution on this area will probably be crippled to half of it (in the green channel). So much for the usability of real (Schneider) or short retrofocus (Rodenstock) wideangles.

It is simply not logical.

Regards
Stefan

 

[DougDolde]

and: stitching for XXL file size/resolution
and: a tech cam lens will out-resolve a MF dSLR lens

I compared the Phase One -Schneider 55mm with the Roddy 60mm, and it was very easy to see how much better the tech lens was (everywhere but the very center, where it was more evenly matched). If you want to walk around doing landscapes with some normal wide (40-60mm) you are better off with a tech lens' resolving power than one designed for your DF/XF camera.

(I don't think its that Schneider/Phase make bad lenses, but that it's a function of strong retrofocus vs mild retrofocus)

Yes but the new P1 Schneider 40-80mm and the 35mm are amazingly sharp

 

[voidshatter]

After seeing some of the horrible posted images I thought I would at least test my Leaf/Credo 50 to the limits. I know.... its not the 100 but its a Sony sensor and probably the basis for the 100. I used my 35mm XL and added rise of 15mm...... that's 15mm folks on the this legendary love/hate lens. Please forgive any sensor dirt..... I think that Phase will have cast issues under control. I would never, ever shift my 35mm XL 15mm but this is just an example of cast with CMOS and the Phase fix.

Victor

At the first glance I said holy sh1t this is awesome and I would like to replace my 40HR with the 35XL for portability! On a second thought I noticed that you shifted along the longer edge of the sensor (i.e. you shifted upwards in portrait direction), which is less stressful for a Sony CMOS sensor. If you shift along the shorter edge of the sensor (e.g. shift upwards in landscape direction) I strongly doubt that it can hold this well.

 

[torger]

At the first glance I said holy sh1t this is awesome and I would like to replace my 40HR with the 35XL for portability! On a second thought I noticed that you shifted along the longer edge of the sensor (i.e. you shifted upwards in portrait direction), which is less stressful for a Sony CMOS sensor. If you shift along the shorter edge of the sensor (e.g. shift upwards in landscape direction) I strongly doubt that it can hold this well.

Low saturation colors also make things look less bad than it is.

It's incredible how patient we tech cam users are, we force pieces together that obviously was not designed for eachother. Schneider Digitar wide on a Sony sensor is like putting truck tyres on a formula one car, indeed two high end things but not made for eachother at all. We rely heavily on the over-capacity of the incredible Sony sensor, over-capacity on DR to compensate the gross signal loss due to pixel shading, and over-capacity in tonality to compensate the gross loss in that due to crosstalk.

With the IQ3100 and Rodenstock lenses it looks better, very much better it seems. With voidshatter's recent tests I'm much more positive than before, but make no mistake we're doing the same thing here just to a lesser extent -- we rely on Sony's over-capacity in tonality and DR to compensate the fact that the Rodenstock lenses were not designed for this sensor type. Or we shoot center frames, and I think the A-series will be 100% fantastic with the IQ3100.

 

[vjbelle]

At the first glance I said holy sh1t this is awesome and I would like to replace my 40HR with the 35XL for portability! On a second thought I noticed that you shifted along the longer edge of the sensor (i.e. you shifted upwards in portrait direction), which is less stressful for a Sony CMOS sensor. If you shift along the shorter edge of the sensor (e.g. shift upwards in landscape direction) I strongly doubt that it can hold this well.

These images are shifted in landscape position 10mm and 15mm with LCC's before and after corrections. Again, this is a crop 44X33 Sony Sensor but I have confidence that Phase can correct any issues with color shift for the 100MP at least for my mighty 35mm XL. Processed in C1 9.0.2.13. I will never sell this lens!!

Victor

 

[voidshatter]

These images are shifted in landscape position 10mm and 15mm with LCC's before and after corrections. Again, this is a crop 44X33 Sony Sensor but I have confidence that Phase can correct any issues with color shift for the 100MP at least for my mighty 35mm XL. Processed in C1 9.0.2.13. I will never sell this lens!!

Victor

Thanks! Do you have some examples of this but with blue sky?

The worst case would be white walls indoors (mazing artifact on the blue part of the LCC file).

 

[vjbelle]

No blue skies today in this neck of the woods...... Even if there were some slight artifacts at 15mm I really do think that Phase has worked hard on casting issues and given the bleak forecast regarding LF lenses I think that this is all going to have to be handled on the software side. They've shown, to me, that they can with the 35XL...... I get the best of many worlds - low distortion, low casting, low weight. I compare this to the giant Rody 32mm and its tremendous out of pocket expense and I'm thankful I never succumbed to the urge to click the 'add to basket' button.

Victor

 

[jlm]

sort of a second tier question, but what about the other IQ3 models advertised, like the IQ3-50? are they in production or just drawing board?

 

[jagsiva]

sort of a second tier question, but what about the other IQ3 models advertised, like the IQ3-50? are they in production or just drawing board?

The other IQ3 backs were released a few months ago:

IQ3-50 same sensor as IQ250 CMOS
IQ3-80 same sensor as IQ280 CCD (but now has long exposure similar to IQ260)
There is no IQ3-60 (correction - there is a 60, no IQ3-40).

The new back is the IQ3-100MP. So really an extension to the line. So IQ3 has 50/100MP CMOS backs and 80MP CCD back.

Hope this helps.

 

[torger]

I haven't used the latest C1. Anyone who knows if they have made any real improvements of lcc correction for the Sony 50MP since the IQ250 release? Maybe better at avoiding mazing and desaturation than before?

(Completely reversing crosstalk is a very difficult problem, but possibly you could hide some of its effects with some workarounds. The SK35 raws I saw for the IQ250 when it was new was pretty awful)

The extreme DR and low noise does give a better base than ever to make some pretty advanced signal processing tricks.